GPT 5.4 is now available in Cursor! We've found it to be more natural and assertive than previous models. It's currently the leader on our internal benchmarks.

@cursor_ai 👀

@cursor_ai 🙌

@cursor_ai Cant believe this was announced before the official OpenAI announcement I gtg try it now!!!

Pick which twin will win! You could score a share of $2 million in Bonus Bets.

@cursor_ai I like my model assertive

@cursor_ai Where is GPT 5.4 Pro? The old OpenAI Pro models are gone.

@cursor_ai natural and assertive gpt 5.4 is just a senior dev who’s tired of seeing me use 'any' in typescript. finally, an llm that will roast my code before i even hit save

@cursor_ai Which model was the previous leader? Opus or 5.3 codex?

@cursor_ai Excited to see how teams use this to build smarter tools

@cursor_ai dude that sounds like a total game changer I'm lowkey excited to see where this goes from here

@cursor_ai Yes!

@cursor_ai is this better than opus 4.6 and codex 5.3 then?

Free Cone Day is here! Come celebrate with us at a DQ location on today by getting a FREE small vanilla cone 🍦

@cursor_ai Shoutout Gpt5.4

@cursor_ai how much?

@cursor_ai Lfg

@cursor_ai WHAT THE FUCK

@cursor_ai That was fast, thanks 🫡

@cursor_ai Why MAX only though @leerob?

It’s time to dance! Get it on tournament action with Bonus Bets from FanDuel.

@cursor_ai The entire squad is MAX Only.

@cursor_ai "So it's not called Codex anymore, but it's better than codex for coding" "Yes in the past codex models were better at coding than non codex, but we decided to change the naming" Stay tuned for garlic, mini, max and AGI plus!

@cursor_ai Within a minute you are ready with new model?

@cursor_ai But how fast it eats tokens compared to sonnet 4.6?

@cursor_ai Let's gooo

@cursor_ai What's the difference between using Cursor vs Codex? Both have direct access to GPT 5.4

@cursor_ai 🐐 cursor

@cursor_ai @JuanRezzio Should I use Codex or GPT 5.4? So confusing.

Full guide for startup builders

@cursor_ai Love it. The real test for me is: fewer regressions + better multi-file changes under tight guardrails. Excited to run it through some gnarly repos. 👍

@cursor_ai "more assertive" is interesting, most models go the other way to avoid pushback

@cursor_ai How does the "assertive" part show up in practice? Are you seeing it push back on requests or just be more confident in its suggestions?

@cursor_ai Team was ready to share this from the announcement tweet

@cursor_ai i was really excited

@cursor_ai i'm not a begger, but pleeease share the benchmarks 😏

@cursor_ai that was fast

@cursor_ai By assertive, do you mean it will shout back at us when we shout at it?